Mapping the medullar adiposity of lumbar spine in MRI: A feasibility study

PURPOSE: The bone medullar adiposity is a marker of bone quality to the point that there is a need to investigate the factors which influence or not the density and distribution of this fat in the spine, especially at the lumbar level. The purpose was to test the feasibility of a Dixon three-point technique and investigate the vertebral marrow fat distribution. MATERIAL AND METHODS: A sagittal sequence Iterative Decomposition of Water and Fat with Echo Asymmetry and Least-squares Estimation (IDEAL) IQ was performed on the lumbar spine of 46 subjects who were not suffering from any bone disease (21 women and 25 men, aged 18–77 years). Medulla adiposity was determined directly from the measurement of the fat fraction in each vertebral body (T12 to S1) obtained on the fat cartography automatically generated by the IDEAL sequence. RESULTS: Average vertebral fat fraction was 36.48% (SD 12.82), with a tendency to increase with age and to higher values among men. We observed a craniocaudal gradient of the fat fraction (β = 1.37; p < 0,001; SD 0.11) increasing with age in the lumbar spine from T12 to L5. Through multivariate analysis, this gradient was adjusted for sex, weight and height of the subjects. CONCLUSION: This feasibility study shows the existence of a physiological craniocaudal gradient of vertebral medullar adiposity from T12 to L5. This gradient increases with age but it is independent of sex or BMI. The IDEAL sequence allows quick and reproducible measurement of the spine vertebral medullar adiposity.